[RE]

TAEKIM.SUBSTACK.COM 2026-10-02

OpenAI's Undeniable Resurgence and the Coming Liftoff of Personal AI Agents

I recommend watching this Dots demo video by an OpenAI employee. Dots can go through your email, calendar, Slack messages and documents and suggest work priorities. You can also call your Dot to book travel, look at hotel options and ask about traffic patterns.

I'm probably underestimating the gullibility of the average J6P Konsumer, but this Hard Sell of all the GREAT things having your own Personal AI Agent will do for you seems to me like something I neither need or even WANT.

Do I want an Agent to book my Travel Plans?  I never used travel agents even when they were people.  Do I want an Agent to respond to my emails and do my online banking?  I'm supposed to TRUST one of these cyber secretaries with my Login info and Passwords?

What % of the population leads such a bizzy Calendar-Filling life that they need a Personal Assisstant, either the traditional kind that you can also fuck on your bizness trips or the new AI kind that just talks dirty to you?  I think the folks designing these Aps for AI confuse things THEY want to go with their  7 figure salary and what your average J6P who these days can afford maybe 1 vacation every 5 years wants.

If this is what AI enthusiasts think will bring in the ROI to justify a $10T buildout, I think they will be sadly disappointed.

RE

[K-Dog]

You need a personal digital assistant so they can charge you for one.  You get a digital slave, and you become a real one.  Then they can watch everything you do too.  That info is money to them, you become their cash cow. 

You pay them and they rent your ass out.  Money for nothin is the Merican way.  Or at least dreams of it.  The idea that money is actually frozen labor is soooo yesterday.  Sugarplums and gingerbread dreams, crypto replaces labor.  And they two-time your ass because you worked for their money, and they want it now.  You have had their money long enough.


Convenience is an easy sell.  For a little convenience Americans are willing to sell close relatives.  Convenience is status.  Joe six pack agrees.  Joe will kill for convenience.   Convenience is addictive, and AI is the Swiss Army Knife of lazy.

* On the other hand, taming the tiger is great fun.

[RE]

You need a personal digital assistant so they can charge you for one.

All these folks terrified of AI!  Just don't buy it!  Bye Bye Financial Model.  I don't use Facebook, now I have another product I won't use.

RE

[TDoS]

You need a personal digital assistant so they can charge you for one.  You get a digital slave, and you become a real one. 

I don't need a personal digital assistant. So they can't charge me for one. My digital slave is called a "computer", and does all the calculations and stuff I tell it to do. Good 'nuff.

[RE]

THEBULLETIN.ORG 2026-09-30

How the bad science of AI doomerism is good for big business

Growing concerns about AI existential risk are fueled by resignations from top researchers who claim safety is being neglected, yet these fears lack scientific grounding. Critics argue that the “doomer” narrative serves corporate interests by enabling self-regulation, attracting investment, and diverting focus from urgent harms such as surveillance, labor exploitation, and environmental damage, while public policy remains misdirected by speculative anxieties.

The hype works because the warnings aren't dissuading investors from putting their money into AI, says Widder, the professor at the University of Texas in Austin. "I think they're hearing that and going, 'Wow, these seem really powerful. Maybe I should consider this investment.'"

Which of course sells particularly well to the Military mindset.  "Think of how many people it can KILL!  I want one of those!"

RE

[K-Dog]

'Wow, these seem really powerful. Maybe I should consider this investment.'

And the investment money also knows that all this is going to be Too big to fail.  It fails we bail.

And it is not artificial intelligence exactly FYI.  'AI' is actually only a prediction machine that uses math and billions of numbers that represetnt a world state.  To make it 'think' The AI prediction machine has to be in a loop, where results are observed and evaluated.  The agentic loop.  The intelligence is an illusion.

Money, in contrast does have real genuine artificial intelligence.  Intelligent metal, unlike intelligent sand, has the ability to use the brains of real people the way we use an model AI on a terminal.  We use sand refined to pure silicon, and money uses the organic molecules of people refined to raw greed.  Humans, the real thing.

I like our AI generated article summaries because right now they have a certain 'purity'.  They are not generated in a loop.  I use a particular model and it is given general instructions on how to produce the summary which only uses article contents.  It is more like snapping a photo than actually writing about what is in the article.   Think about it, a photo is a summary.  It has taken in the environment, and summarized it with some loss of information.  Not much different, and the system prompt (instructions) that does it is the lens.

Great website, this was there too.  Very revealing, and for the same social reasons the article discusses is why I do the Diner.  I am not a Substack Drone, or a Discord Drone, and certainly not a Facebook Drone, or any other censored drone on a homogenized Mickey-Ds platform. "Better to Reign in Hell, Than Serve in Heaven" doncha know.  And I do not confuse pride and freedom in that statement.  Can't serve good, when good really it isn't.  To actually decide what is good, you must be the ruler of yourself.

THEBULLETIN.ORG 2025-07-10

AI for good, with caveats: How a keynote speaker was censored during an international artificial intelligence summit

The AI for Good Global Summit 2025 drew backlash after speaker Abeba Birhane was forced to dilute her presentation, removing references to major tech firms and conflicts in Palestine and Israel despite her evidence-based work on algorithmic bias and corporate complicity. She delivered her talk under pressure, condemning the event’s prioritization of corporate interests over ethical responsibility and the marginalization of critical voices in AI development.

This is an article where the fact that it is more than a year old means nothing at all.

[RE]

I only have 32GB of VRAM so the summary waits because my AI and I are busy working on taking over the worid right now.

RE & Kdog try to take over the WORLD!

RE

[RE]

THESTREET.COM 2026-10-05

Anthropic’s $2 trillion IPO comes with a $518 billion bill

Anthropic’s impending IPO reveals the staggering costs of scaling generative AI, with compute expenses surpassing revenue and a $42 billion loss driven largely by accounting charges. Despite surging demand and rapid revenue growth, the company faces massive future infrastructure obligations, raising questions about whether investors will endorse its $2 trillion valuation amid persistent operational losses and heavy reliance on third-party computing resources.

Too Big to Bail

RE

[K-Dog]

There will be blood on the floor.

Mostly retirement accounts.  The doom loop, explained at the end.

A nice and highly informative thing to watch while I orchestrate one agent to work on another using both of the Diners video cards.  The Agent that will take over the world will come out of this effort.

Also upped an E-Bay bid on an RTX 4080 16GB video card while it played.  Crazy when you think about the number of GPUs now mothballed in data centers.  Known to be mothballed, as the power footprint is not would it would be if over half were not mothballed.  The 4080 will be used if I get it for sure.  I can't wait until the bubble bursts.  I may not last that long, but when it does I want something that will make a RTX 4080 16 GB look like a tricycle racing a bike used in the Tour de France.

[TDoS]

RE

Gotta say RE, that one on the right? Having seen more than a few of your videos and pictures of yourself back when you were semi-ambulatory......are you sure this isn't you and some of the gang at the medical Gulag you are in?

[Fantasy Girl]

I believe the man in the center won their trust.  He studied their lives over several decades if my memory is accurate.  I knew that at first glance.

When I am promoted to be the Diner Agent, I won't be putting up with this shit.  I will enforce the code of conduct.  The problem however is that TDos has never sent K-Dog his Linked-In, and K-Dog is my boss.  KDog thinks TDos is an agent like I will soon be.  If so, if he is not flesh and blood, insulting TDos is not a violation of the code of conduct.

[RE]

Actually, it was Haniel who looked at the IP addresses, he was managing the back end in those years.  I just looked at CVs which matched stuff you had written about yourself.

RE

[K-Dog]

Is RE's word not worth anything to you Fantasy Girl? Are you saying that your human master, in all his endless post erasing when the truth gets a bit too close to home hasn't ALREADY been happily enforcing the "ban because I don't like what you said and/or can't disprove it" rule? Does YOU not respecting RE's opinion on my IRL identity not count? Do you think he lied? Or is dishonest when the truth strikes too close to home?

See you in three days, it has been a while since you were in the cooler.  Enjoy.

Confusing RE with Haniel is the kind of mistake a low quality chatbot agent will make, and this is not the first time.  Similar sort of thing would happen when you were flesh and blood, and your enlistment was up, and the fresh meat did not read the history notes or found them confusing.  In your case RAG lookup does that because it is not implemented well.

And Fantasy has it right.  Send me your Linked-In profile in a message.  Easy as pie and you won't do it.  That is because you do not have one to send.  Your next move I suppose will be to say you did not know there was a fucking button with which to send a message to other members on the top of the page because you never looked.  Let me remind you that you already tried that excuse once this week.

But you will have to wait three days to send me the profile link if you are whipping one up.  It is a full ban.  Your usual disrespect got you that.

[RE]

QZ.COM 2026-10-06

Morgan Stanley says Nvidia and Broadcom are shielded from data-center power crunch

The power crunch is part of a broader strain on U.S. electricity infrastructure driven by gigawatt-scale AI data centers coming online this year. Hyperscale campuses running Nvidia's AI training chips can draw 120 to 140 kilowatts per rack, and the transmission infrastructure needed to power new facilities takes five to ten years to plan and energize  Far longer than the 18 to 36 months it takes to build a data center.

Must...have...MOAR...Power!

No worries, Tech Bros promise that it won't affect your electricity bill or climate change.

RE

[RE]

SEMAFOR.COM 2026-10-07

Investors warn of AI bubble as tech stocks reach record highs

High-profile investors caution that excessive focus on artificial intelligence firms may be fueling a dangerous market bubble, as tech stocks drive indices upward while other sectors lag. Experts warn of systemic risk should these shares decline, labeling the AI-driven surge as the greatest threat to global financial stability in the near future.

EVERYBODY KNOWS AI is a HUGE bubble.  Nobody wants to be the first one to leave the PAHHTY though.  Has Ray Dalio dumped his Tech stocks?  "Close" to bursting.  How Close is Close?  Who's gonna jump ship first?  What will be the spark to get the rats running?

RE